Modeling Systematicity and Individuality in Nonlinear Second Language Development: The Case of English Grammatical Morphemes

Abstract

This article introduces two sophisticated statistical modeling techniques that allow researchers to analyze systematicity, individual variation, and nonlinearity in second language (L2) development. Generalized linear mixed-effects models can be used to quantify individual variation and examine systematic effects simultaneously, and generalized additive mixed models allow for the examination of systematicity, individuality, and nonlinearity within a single model. Based on a longitudinal learner corpus, this article illustrates the usefulness of these models in the context of L2 accuracy development of English grammatical morphemes. I discuss the strengths of each technique and the ways in which these techniques can benefit L2 acquisition research, further highlighting the importance of accounting for individual variation in modeling L2 development.

Multi‐Word Expressions in Second Language Writing: A Large‐Scale Longitudinal Learner Corpus Study

“…Mixed‐effects models present various advantages compared to means‐based parametric statistical techniques such as ANOVAs and t tests (Cunnings, ; Cunnings & Finlayson, ; Gries, ; Linck & Cunnings, ; Murakami, ). For example, as Cunnings () noted, the fixed effects component in a model can comfortably accommodate multiple independent variables, including categorical variables (males vs. females, L1 vs. L2), continuous predictors (age), as well as a mixture of both.…”

“…Ironically, Gries () dubbed mixed‐effects models “the most under‐used statistical method in corpus linguistics” (p. 95), and, by extension, in learner corpus research. Only a handful of published studies have employed mixed‐effects modeling for (various) longitudinal learner corpus data (Crossley, Skalicky, Kyle, & Monteiro, ; Crosthwaite & Jiang, ; Meunier & Littré, ; Murakami, ) and only one study to date has used mixed‐effects models to explore L2 phraseological development over a period of time (Spina, , see above). Echoing Cunnings (), Gries () noted that corpus linguistics can benefit from mixed‐effects models for the reasons similar to those for which psycholinguists have been using them for over a decade.…”

“…A strong argument has been advanced for the inadequacy of the currently used means‐based statistical techniques such as ANOVAs and t tests. On the contrary, mixed‐effects models have been shown to be much better suited for the analysis of a variety of data types including corpus data (Gries, ; Murakami, ).…”

Using Explicit Instruction About L1 to Reduce Crosslinguistic Effects in L2 Grammar Learning: Evidence From Oral Production in L2 French

“…A MAJOR FOCUS OF SECOND LANGUAGE acquisition (SLA) research to date has sought to understand the competition and relationships between a learner's different languages (Calabria et al., ). This research has repeatedly shown that the use of a single language activates a speaker's other known languages (Marian & Spivey, ; Wu & Thierry, ), that prior first language (L1) knowledge and experience can influence second language (L2) use (e.g., selective attention to linguistic cues; Ellis & Sagarra, ; MacWhinney, ), and that L1–L2 differences can influence the route and rate of L2 morphosyntactic development and processing (Avery & Marsden, 2019; Isabelli, ; McManus, , ; Murakami, ; Roberts & Liszka, ). However, despite major advances in what we know about the cognitive effects and mechanisms of learning a second language, little research has systematically examined the next step in this program: How can this understanding about the competition and relationships among a learner's different languages be used to facilitate L2 learning and teaching?…”

Individual developmental trajectories in the L2 acquisition of Spanish spirantization

“…Because mixed-effects models have been described extensively elsewhere (Cunnings & Finlayson, 2015;Curran, Obeidat, & Lisardo, 2010;Linck & Cunnings, 2015;Murakami, 2016;Singer & Willet, 2003), only a brief overview is presented here. Growth curve models are a subclass of mixed-effects models that estimate between-and within-individual change over time through fixed and random effects.…”